Jeff Brandt, Editor of PinHawk's Law Technology Digest Newsletter, discusses on how Bruce Schneier's art of bypassing AI security is direct and to the point.
Bruce Schneier is direct and to the point. "Researchers have demonstrated that putting words in ASCII art can cause LLMs-GPT-3.5, GPT-4, Gemini, Claude, and Llama2-to ignore their safety instructions." That's crazy! Read more at Schneier on Security: Jailbreaking LLMs with ASCII Art
Published March 15, 2024.
